Citation | Study group | Study type (level of evidence) | Outcome(s) | Key result | Comments |
Wang et al 6 | 118 children aged 3–17 with at least 4-week history of at least weekly ‘moderate‐to‐severe headache with a throbbing or pulsatile quality, associated anorexia/nausea, vomiting, photophobia, sonophobia, or relief with sleep, no fever or evidence of infection.’ Given either MgO (9 mg/kg per day 3×/day) or placebo for 12–16 weeks. | Randomised, double-blind, placebo-controlled, parallel- group trial (level 2b) | Number of days with at least one headache during each of eight 2-week intervals. | 86 (73%) completed the study: 44 in the placebo group, 42 receiving MgO. Intention-to-treat analysis showed a statistically significant reduction in the number of headaches over time in the treatment group (p=0.0037), but not significant for a trend over time in the placebo group (p=0.086). Reduction in severity of the headache in the treatment group (p=0.0029). Patients taking Mg reported more soft stools/diarrhoea (19% vs 7%) (p=0.04). | Large variability between the participants. High level of attrition. Those who had chronic daily headache were not identified. No power calculation. Plasma Mg levels during the trial were not monitored. |
Gallelli et al 7 | 160 participants aged 5–16 years old, 80 M (mean age 10.64±2.29), 80 F (mean age 12.15±2.66) with at least four attacks per month. Divided into 4 groups: treatment with paracetamol (15 mg/kg) or ibuprofen (10 mg/kg) without or with Mg (400 mg/daily). Results evaluated every month. Treatment given for 18 months. Groups were non-randomly assigned based on age and gender. | Single-blind, balanced-recruitment, parallel-group study (level 2b) | Timing of pain relief and intensity differences from baseline of 0–3 hours after treatment. Effect treatment had on exacerbation of migraine and any adverse drug reactions. | Number of episodes and pain endurance similar in all four groups. Paracetamol and ibuprofen caused reduction in pain intensity (p<0.01), with no time-dependent correlation. The reduction of acute pain was faster with ibuprofen than with paracetamol (p=0.004, 95% CI −27.54 to −5.54). The frequency of migraines was not modified. If given Mg pretreatment, in both the paracetamol and ibuprofen groups, then there was a significant reduction in the intensity of pain (p<0.01) with no time-dependent correlation, and significantly reduced pain frequency (p<0.01). With Mg pretreatment there was also a significant reduction in pain relief timing with the paracetamol treatment (p=0.022; 95% CI 1.99 to 24.17) but not with the ibuprofen treatment (p=0.737; 95% CI −4.74 to 6.64). Results not age related. The efficacy of ibuprofen and paracetamol was increased with Mg with no age-related effects. No adverse drug reactions recorded in either the paracetamol or ibuprofen groups when in association with magnesium. | No placebo groups. Treatment was only administered over a short period of time. None of the participants had chronic migraine. The study was open label with no randomisation. |
Kovacevic et al 9 | 34 children,15 M, 17 F (two males lost to follow up), aged 7–17 years (mean 12.44 (2.54)) with migraine treated with oral prophylaxis Mg (4–6 mg/kg per day) over a period of 6 months. | Follow-up cohort study (level 2b) | The change in, quality of life (assessed by KIDSCREEN-27), disability levels (assessed by PedMIDAS)and depressive and anxiety symptoms (assessed by RCADS) | The score from PedMIDAS (score used to assess migraine disability) was significantly reduced from baseline to the end of the study (p<0.001). The KIDSCREEN-27 score, which looks at the psychological and physical well-being and social support, significantly increased from baseline to the end of the study; physical well-being (p=0.01); psychosocial well-being (p<0.001). The RCADS score for depressive symptoms significantly decreased from baseline to endpoint (p=0.001) and for anxiety symptoms (p=0.001). No statistically significant changes in KIDSCREEN or RCADS parent-report scores. Reduction in the frequency of migraine attacks after the 6 months of treatment (p≤0.01). | Potential selection bias as participants selected from a tertiary centre. Different forms of Mg used throughout the study which may have had different bioavailability (MgO vs Mg glycinate) No patient blinding—may have influenced questionnaire responses. No assessment of plasma Mg concentration before and after treatment was noted. |
Moscano et al 10 | A fixed dose of the supplement Partena tablet (Andrographis paniculata, CoQ10, riboflavin and Mg) given for 16 weeks to 91 children with migraine with aura, without aura or tension-type headaches. Assessed at week 0 (T0), 8 (T1), end of treatment window of 16 weeks (T2) and the follow-up period of week 20 and week 32. | Observational multicentre study (level 4) | Frequency of headache and the pain intensity of headache episodes. | Number of headache episodes fell in the group with migraine with aura during treatment (p<0.01) and after removal of the treatment (p<0.01). In patients with migraine without aura there was a fall in the number of headache episodes throughout treatment (p<0.01) although not when the study had ended (p>0.05). Reduction in pain intensity in the group with migraine with aura throughout treatment (p<0.001) and also after treatment had ended (p<0.001). Same effects seen in patients with migraine without aura during treatment (p<0.001) and after treatment had ended (p<0.001). | Small sample size. No placebo and no control group. Short observational period. |
CoQ10, coenzyme Q10; F, female; M, male; Mg, magnesium; MgO, magnesium oxide; PedMIDAS, Pediatric Migraine Disability Assessment Scale; RCADS, Revised Child Anxiety and Depression Scale.